Transbronchial Cryobiopsy in Interstitial Lung Disease

Safety of a Standardized Procedure

Abstract

Background: 

Transbronchial lung cryobiopsy (TBCB) plays an increasing role in the evaluation of diffuse parenchymal lung disease with acceptable diagnostic yield and safety profile compared with that of video-assisted thoracoscopic surgical lung biopsy on most reports. However, published outcomes with regard to safety and complication rates vary. We aim to determine the safety profile of TBCB when performed using a standardized protocol consistent with recently published expert guidelines.

Materials and Methods: 

We reviewed prospectively maintained databases from 5 Australian tertiary referral centers. The procedures were performed in accordance with a recent expert statement recommending standardization of TBCB procedures, in particular with fluoroscopy, a secured airway, and prophylactic bronchial blockers. Periprocedural complications were assessed along with clinical outcomes.

Results: 

A total of 121 patients underwent TBCB between August 2013 and August 2017 following a standardized protocol using general anesthesia. Of them, 84 patients (66.7%) were discharged on the day of the procedure. Pneumothorax occurred in 18 patients (14.9%), 13 (10.7%) of whom required chest tube drainage. Moderate bleeding occurred in 15 patients (13.2%) and severe bleeding in 1 (0.83%). Histopathologic diagnosis was made in 80 patients (66.1%).

Conclusion: 

Pooled outcomes from Australian tertiary centers indicate that TBCB is safe when performed in a protocolized fashion. Active measures to anticipate and manage bleeding and to direct biopsy position result in low rates of major complications.

Transbronchial cryobiopsy (TBCB) is increasingly used in the assessment of diffuse parenchymal lung disease, with a recent meta-analysis showing a diagnostic yield of 83.7% compared with 92.7% obtained by video-assisted thoracoscopic surgery.¹ Significant variability in procedural safety outcomes exists in published reports.² Although most publications have reported low rates of pneumothorax and significant bleeding, isolated reports have described higher rates of complications.³ Factors contributing to higher rates of complications have been recognized, and recent calls for procedural standardization have been followed by the publication of an expert statement on safety and utility.⁴

Cryobiopsy was independently implemented at 5 tertiary centers in Australia. Performance of procedures was consistent with recently published guidelines with regard to optimal technique of TBCB. We present this multicenter cohort with the aim of describing procedural outcomes when TBCB is performed using a standardized technique independently across multiple sites.

MATERIALS AND METHODS

A review of prospectively maintained databases of TBCB procedures was performed at 5 referral hospitals in Australia, with all data handled anonymously and according to institutional ethics requirements. Hospital files and operation reports, lung function results, and pathology reports were sourced for relevant information.

Hospital files were used to obtain baseline characteristics of patients, including age, sex, anticoagulation and/or antiplatelet therapy, and pulmonary function. Procedure reports were used to obtain procedural details, including the use of general anesthesia, rigid bronchoscopy, number of biopsies, number of lobes, and lung segments biopsied. Other aspects including perioperative complications, length of hospital stay, acute exacerbation of interstitial lung disease (ILD) within 7 days of the procedure, and 30-day mortality were also assessed. Pathology reports, multidisciplinary meeting reports, and treating physicians’ notes were reviewed.

Performance of Cryobiopsy

A procedural protocol was developed by Australian interventional pulmonologists performing transbronchial cryobiopsies to standardize the procedure. This protocol is in accordance with the recently published expert statement to reduce complications.⁴ The statement indicates that TBCBs should be performed in intubated patients with deep sedation or general anesthesia, with use of prophylactic balloon blocker (Ardnt bronchial blockers; Cook Medical, Bloomington, IN or Fogarty Occlusion catheters; Edwards Lifesciences, Irvine, CA were used), with fluoroscopic guidance, and performed in the operating room with full anesthesia support or in a dedicated bronchoscopy suite with emergency equipment immediately available.⁴ Anticoagulants and antiplatelets were withheld at the time of biopsy when possible.

Definition of Complications

Our definition of bleeding closely aligns with previously described classifications, with minor changes due to the prophylactic use of a balloon blocker.⁵ Mild bleeding was defined as any bleeding that did not alter the course of the procedure and only required routine measures of bronchoscopic occlusion/suction/tamponade, cold saline, topical adrenaline, topical tranexamic acid, and prophylactic balloon tamponade. Moderate bleeding was defined as any intraoperative bleeding requiring nonroutine measures such as the use of an extra or a more proximal endobronchial balloon to control bleeding. If bleeding had affected the course of the procedure, by significantly prolonging the procedure or by limiting the biopsy numbers or locations, it was categorized as moderate as well. Severe bleeding was defined as any bleeding causing significant clinical instability including hypoxemia or hypotension, and/or requiring further aggressive measures to stabilize the patient, such as intubation and prolonged ventilation, double-lumen intubation, blood transfusion, parenteral tranexamic acid, or other coagulating factors.

Assessment of Clinical Outcomes

All histopathology reports were reviewed and examined by a single investigator. Final clinical diagnoses were sourced from treating physicians’ notes.

RESULTS

A total of 126 patients underwent TBCB for the evaluation of interstitial lung infiltrates at 5 referral hospitals in New South Wales and Victoria, Australia, from August 2013 to August 2017. Of these 126 procedures, 5 initial cases from a single center were excluded, as they were performed before introduction of the standardized protocol.

Patient characteristics of the 121 included cases are listed in Table 1. All centers used ERBOKRYO CA or ERBE 2.0 CRYO equipment (Erbe, Solingen, Germany); however, probe size selection (1.9 vs. 2.4 mm) and freezing time (from 3 to 5 s) varied according to the proceduralist’s preference. Fluoroscopy for optimal cryoprobe placement, generally 1 to 2 cm from the visceral pleura, was used in all cases. All procedures were performed using rigid bronchoscopy or endotracheal tube and under general anesthesia. Further procedural details are summarized in Table 2.

TABLE 1: Baseline Characteristics of Patients Undergoing Transbronchial Lung Cryobiopsy for the Diagnosis of Diffuse Parenchymal Lung Disease

TABLE 2: Characteristics of Transbronchial Lung Cryobiopsy Procedures for the Diagnosis of Diffuse Parenchymal Lung Disease

Safety

Complications are listed in Table 3. TBCB was generally safe with none or only mild bleeding occurring in 105 cases (86.8%). Pneumothorax occurred in 18 patients (14.9%), with 13 requiring an intercostal catheter (ICC) (10.7%). Over two thirds (84; 69.4%) of patients were discharged on the same day, and a further 9 (7.4%) were discharged the next day (Table 3). There was 1 case of severe bleeding that occurred under general anesthesia with rigid bronchoscopy and prophylactic balloon blockade. In this case, the balloon was dislodged during the procedure and required insertion of a double-lumen ETT, mechanical ventilation for 24 hours, and blood transfusion. The patient subsequently made a full recovery to baseline function.

TABLE 3: Complications and Length of Hospital stay for Transbronchial Lung Cryobiopsy Procedures

There was only 1 case of ILD exacerbation within 7 days of the procedure in a patient with final diagnosis of cellular nonspecific interstitial pneumonia (NSIP). This patient, who also suffered from a pneumothorax requiring an ICC and suction, eventually improved after a course of systemic steroids and was discharged. One death was noted (mortality 0.83%), occurring on day 3 after the procedure due to a massive pulmonary embolus. This patient also suffered a pneumothorax requiring ICC and ward admission. Aside from TED stockings, no parenteral deep venous thrombosis prophylaxis was administered. An investigation found this incident was not related to the specific cryobiopsy procedure and led to the recommendation of standard deep venous thrombosis prophylactic guidelines in patients admitted following this procedure.

Clinical Outcomes

TBCB alone yielded a confident histologic diagnosis in 80 patients (66.1%). Histopathologic and eventual clinical diagnoses are listed in Table 4.

TABLE 4: Final Histopathologic and Clinical Diagnoses in Patients With Interstitial Lung Infiltrates

DISCUSSION

We present the first multicenter cohort of outcomes from TBCB performed according to a standardized protocol in ILD diagnosis from Australia. The recently published expert statement⁴ emphasizes the need for standardization of the technique, including the use of fluoroscopy, a secure airway, and balloon blockade as bleeding prophylaxis. Our experience demonstrates the importance of adherence to this recommended protocol for maintaining procedural safety, with rates of bleeding (moderate bleeding 12.4%, severe bleeding 0.8%) and pneumothorax (14.9%, ICC in 10.7%) lower than in many prior studies, particularly those wherein a nonstandardized protocol was used, or fluoroscopy/prophylactic balloon tamponade was not universally utilized.³ In our cohort, when TBCB was performed using a standardized protocol adhering to expert guidelines, an acceptable safety profile was demonstrated.

A histopathologic diagnosis was reached in 66.1%, with the most common final diagnoses of NSIP and hypersensitivity pneumonitis. In our cohort, we included all TBCBs performed for the diagnosis of any diffuse interstitial lung abnormality that eventually yielded almost 10% malignancy. The prevalence of usual interstitial pneumonia or idiopathic pulmonary fibrosis in our cohort at 9.9% is modestly lower compared with some other reports.^6,7 These 2 factors should be considered while comparing our clinical outcomes with other reports (Table 5).

TABLE 5: Comparison of Similar Cohorts Undergoing Transbronchial Cryobiopsy for the Diagnosis of Diffuse Parenchymal Lung Disease

We refined the definition of moderate and severe bleeding considering the fact that bronchial balloon tamponade was used prophylactically in all cases, in accordance with the standardized protocol.⁴ The single case of severe bleeding that occurred was related to the dislodgement of the endobronchial balloon, perhaps emphasizing the value of prophylactic balloon placement. This event occurred early on in the cohort, leading to further training and procedural technique alteration, specifically with regard to endobronchial balloon placement. No further cases of severe bleeding were observed.

Overall the risk of severe bleeding has been reported in the literature to be between 0% and 4% (Table 5). Factors contributing to a reduced risk of bleeding have been identified, and include prophylactic blocker placement and use of fluoroscopy to ensure biopsy is performed in more peripheral areas of the lung. One recent paper from DiBardino et al³ describing outcomes in 25 patients from 1 center illustrated the risks of nonstandardized protocols; they observed bleeding complications in 33% of cases wherein fluoroscopy was not used, and life-threatening bleeding in 1 case wherein cryobiopsy was performed via laryngeal mask airway without a bronchial blocker. Our cohort demonstrates that, in contrast, when performed in accordance with published guidelines, TBCB is a safe procedure.

Possible risk factors for complications in TBCB include the lack of utilization or mismanagement of prophylactic endobronchial balloon blocker, freezing time, and malposition of the cryoprobe at the time of biopsy. As previously reported, bleeding and pneumothorax are generally more prevalent at longer freezing times, that is, 5 or 6 seconds.⁹ In our cohort, freezing times of 3 to 5 seconds were slightly shorter than those of the expert statement guidelines, but this did not have an apparent effect on our outcomes. Other centers that utilize longer freezing times have still produced similar complication profiles.^8,10 The use of a larger probe size of 2.4 mm compared with 1.9 mm is also associated with higher diagnostic yield but a greater rate of pneumothorax.¹

Furthermore, using fluoroscopy to direct the cryoprobe to a suitable distance from the pleura and also away from central and larger vascular structures seems to be an important factor in minimizing complications of pneumothorax and bleeding. This was utilized in all cases in our cohort. Along with the other previously mentioned factors, lack of utilization of fluoroscopy in 40% of cases and heterogeneity in procedural techniques could be responsible for the surprisingly high rate of complications in the DiBardino et al’s³ study. There also seems to be an association between higher complication rates and not using fluoroscopy, both with bleeding and pneumothorax.³

We encountered pneumothoraces in 18 (14.9%) patients, of whom 13 patients (10.7%) required ICC, consistent with previous reports (Table 5). It is unclear whether general anesthesia, used in all of our cases, or jet ventilation, could increase the risk of pneumothorax. In 1 study,⁶ wherein all procedures were performed under deep sedation, pneumothoraces were only encountered in 3% of cases. In contrast, Casoni et al¹⁰ reported a pneumothorax rate of 28% while using rigid intubation. A significant increase in the rate of pneumothorax was also detected when pleura was present in the specimen, compared with when it was absent. These factors likely further highlight the value of probe size and use of fluoroscopy in a standardized protocol to reduce the complication rate.

Patient selection is also an important factor contributing to complication rates. As an example, pulmonary hypertension may contribute to an increased risk of bleeding, and routine echocardiography preoperatively has been suggested in previous literature.^4,10 Because of this, guidelines classify pulmonary hypertension as a relative contraindication.⁴ Despite not routinely assessing for pulmonary hypertension, we observed a low severe bleeding rate. Severely reduced pulmonary function, for example, a diffusing capacity <35% or forced vital capacity <50% has also been suggested as a relative contraindication to TBCB.⁴ In our cohort, 15 (12.4%) patients had a diffusing capacity of the lungs for carbon monoxide <40%, but we did not note differences in the rate of complications in this subgroup.

In our cohort, 84 patients (69.4%) were discharged on the same day, and 9 (7.4%) were discharged the next day, suggesting that TBCB is suitable as an outpatient procedure and that it is well tolerated. The rate of early discharge in our study, however, may be underestimated, as 15 patients in our cohort were either already inpatients at the time of TBCB or stayed >24 hours due to reasons unrelated to the procedure. There was 1 acute exacerbation of ILD within 7 days and 1 mortality due to pulmonary embolism within 30 days (0.83%). Our mortality rate is comparable to similar studies reporting rates of 0% to 1.4% (Table 5). No difference in complication rates was observed between the centers. Surgical lung biopsy, in contrast, is a more invasive procedure requiring intubation and chest tube placement in all patients, with reported 30- and 90-day mortality rates of 2.2% and 3.4%, respectively,¹¹ and reported hospitalization between 4 and 8 days.⁸

Limitations and Future Directions

In our study, all cases were analyzed retrospectively by extracting data from pathology reports, likely producing interobserver variability as one of the major limitations of our study. Clinical diagnoses were not always reached within a multidisciplinary context. Our results, specifically the clinical outcomes, should be interpreted with caution, given the modest prevalence of NSIP (14.9%) and usual interstitial pneumonia (9.9%) in our cohort and the high prevalence of sarcoidosis, lymphoma, or hypersensitivity pneumonitis. Comparability and concordance of histopathologic results between TBCB and video-assisted thoracoscopic surgery remains a matter of debate and is the subject of both recently published¹² and ongoing research.¹³

However, the relatively large, multicenter population can be considered a strength of this study; despite the heterogeneity of the cohort and spread of lung function parameters, the use of a standardized protocol across multiple centers has maintained a low complication rate that can be validly translated to routine clinical practice. In the future, the efficacy of the protocol used may also be further improved through the incorporation of other tools such as radial endobronchial ultrasound^14,15 or computed tomography fluoroscopy¹⁶ to direct TBCB to areas of active abnormality within the lung¹⁷ and away from areas wherein complications are more likely. Future studies are required to determine the additional benefit to diagnostic yield and complication rates when these methods are utilized along with existing, expert-recommended standardized techniques.

CONCLUSIONS

We present a multicenter Australian cohort of TBCB, with all procedures performed in accordance with the recently published expert statement on the safety and utility of TBCB. This protocol is safe and associated with lower rates of complications compared with prior reports utilizing nonstandardized procedures. Our analysis highlights the value of standardizing procedural techniques to minimize complication rates of TBCB.